Abstract:Grounding on the analysis of features of grid based and density based clustering methods, a hybrid clustering algorithm based on grid and density was presented. By clustering in two phases and using only a small number of seed objects in representative units to expand the cluster, the frequency of region query can be decreased, and consequently the cost of time is reduced. An equivalent rule was proposed to make smooth conversion between clustering parameters in that two phases. The algorithm keeps good feature of both density based and grid based clustering methods. It can discover clusters with arbitrary shape with high efficiency and is insensitive to noise. So it is applicable for data mining on large database. The application of the hybrid algorithm in data analysis of accelerometer demonstrates its effectiveness. It is of instructional meaning for the application of data mining in equipment monitoring and faults diagnosis.